She Ran the OS Before She Built the Beat
Before anyone called Kakul Srivastava a music executive, she was the person at a small Canadian startup called Ludicorp who had a side project getting out of hand. The project was Flickr. The year was 2004. She was its first product person, inheriting a user base of 37,000 people who'd stumbled onto something beautiful and chaotic. By the time she left Yahoo! - which had acquired Flickr - she'd scaled that user base past 60 million.
The specific mechanics of how she did it matter: Srivastava was the architect of photo tagging as a consumer concept - a feature so fundamental to the internet that we now forget someone had to invent it. Before Flickr, you couldn't label a person in a photo online. She introduced it first through Adobe's Photoshop Album, years earlier, before scaling the concept to the web at Flickr. Most of what you do on Instagram today traces a line back to decisions she made two decades ago in a Yahoo! conference room.
"In a world where everyone is a creator, Splice is building the essential toolset to make music. Human creativity is our most important resource."- Kakul Srivastava
After Yahoo!, she spent years collecting operating manuals for different kinds of creative platforms. WeWork's Chief Product Officer. GitHub's VP of Product Management - where she made the most significant pricing change in the company's history and pushed open-source development into enterprise. Then back to Adobe to run Creative Cloud, the $9 billion business that powers how designers, photographers, and filmmakers work. Her resume reads like a tour of every major platform for creative people built in the last 25 years.
Splice found her through its board. She joined as a director in 2021, watched the company from the inside, and then in spring 2022 stepped up as CEO - succeeding founder Steve Martocci. It's a rare thing for a board member to become the operator. It means she had already done the due diligence most CEOs never get to do.
Tomfoolery Inc., 2012
Before the corner offices at GitHub and Adobe, there was a tiny startup called Tomfoolery that Srivastava co-founded to build apps for the future of work. Their first product, Anchor, was designed to help teams communicate more openly through "more human interactions." Yahoo! acquired the company in 2014 - years before a podcast app with the same name became ubiquitous. The instinct that built Tomfoolery - that technology should facilitate real human connection, not replace it - runs through everything she builds today.
The Skeptic With the AI Tool
Here's what makes Kakul Srivastava interesting in a world full of tech executives who are, as she puts it, "AI fan girls": she isn't one. She said so out loud, on the record, while running a company that just launched an AI-powered music creation feature. That's either cognitive dissonance or clarity, depending on your priors.
It's clarity. Her argument is surgical: most generative AI music tools are "just not good enough," and the category of push-button music generation - where you describe a mood and a song emerges - is "insulting to musicians." What she's building at Splice with the Create tool is categorically different. It's not replacing the human creative act; it's compressing the tedious parts of it so the human can get to the interesting parts faster.
The "Second Wave" Thesis
Srivastava argues that AI in music is entering its second phase. The first wave was about fear - rights disputes, displacement anxiety, copyright confusion. The second wave, she says, "will be marked by what it gives back: time, powerful new capabilities and expansive creative expression." The preconditions she identifies for this to work: licensed training data, real economic participation for creators, and clear provenance. Without those, it's just extraction.
She cites Daft Punk and Stevie Wonder as canonical examples of what happens when you put disruptive technology in artists' hands. They didn't make robot music. They made transcendent music using machines as instruments. Her evidence for optimism about AI isn't theoretical - it's historical. "Every time we've put technology in the hands of artists, they've surprised us," she said. "Trust the artists."
The clearest proof of her philosophy is Splice's partnership with Universal Music Group to develop ethical AI tools - a deal that explicitly addresses creator compensation and rights. Most music technology companies in 2025 are still arguing about whether to engage with the rights question. Splice already answered it.
"I'm not an AI fan girl. It's a technology that can be used in good ways and really stupid ways."- Kakul Srivastava, on push-button AI in music
Scale is normalized for visualization. Values represent approximate platform metrics during tenure.
Beats Meet Orchestras
In April 2025, Splice announced the acquisition of Spitfire Audio in a deal rumored at around $50 million. Spitfire is the preeminent destination for orchestral and cinematic sample libraries - the tools composers use to score films, TV, and games. Splice is where hip-hop producers find drums and electronic musicians find synth packs. The overlap between these communities was, until 2025, essentially zero.
Srivastava's explanation for the deal was straightforward: "We have such a shared passion for the creators we serve; our values match." But the strategic logic runs deeper. She's been watching a behavioral signal in the creator data - musicians don't stay in one genre anymore. A producer who builds tracks in Splice might be scoring a short film next week. The walls between "electronic music producer" and "composer" have been dissolving for a decade. The acquisition was a bet that the future creator needs a single platform that spans both.
She also told the Financial Times that AI was part of the vision: the combination of Splice's sample library and Spitfire's orchestral depth creates training and tool-building possibilities that neither company could achieve alone. The ambition is a creative platform where a producer can "merge instruments together to get a novel sound that has never been heard before."
Twenty-Five Years of Building Creative Infrastructure
The Record
Named one of the 100 Most Influential People in AI for 2025 by TIME Magazine.
Executives list three consecutive years: 2023, 2024, and 2025.
As GM, scaled from 37,000 users to over 60 million - one of the fastest consumer platform growths of the era.
Led end-to-end experience for the $9B+ Creative Cloud business at Adobe.
Instituted the most significant pricing model change in GitHub's history, driving open-source and enterprise growth.
Named among Top 25 Creative People in Business by Fast Company Magazine in 2016.
Led Splice's ~$50M acquisition of Spitfire Audio, bridging electronic music and orchestral composition.
Introduced consumer photo tagging via Photoshop Album at Adobe - before it became ubiquitous across every social platform.
Kakul on AI, Music and Creative Control
These conversations go deep on Splice's philosophy, the creator economy, and what it actually means to put AI in service of artists.
The Creator Economy's Honest Broker
Most people who run creator platforms talk about empowerment. Srivastava is more precise about what that word actually requires: "Creators need the ability to realize their potential on their own terms with economic ownership and creative control." The qualifier "on their own terms" is doing the work in that sentence. It means the platform's role isn't to own the creative output - it's to facilitate it while leaving the creator in charge of what happens next.
She's also watching cultural signals that most tech executives would dismiss as noise. Vinyl records crossed $1 billion in sales in 2025. Gen Z creators are trading cassettes. Retro hardware is making a comeback. Her read of these trends isn't nostalgia - it's a signal about authenticity. The imperfections and textures of analog creation have human fingerprints on them. That's what people are craving when they buy a record. AI that erases those fingerprints is moving in the wrong direction.
Her formula for what good AI in music looks like: it preserves human momentum rather than replacing the human. Splice's Create tool works by getting a producer to a starting point faster - the rough sketch of a track - so the human can take over and make it their own. The machine opens the door; the artist walks through it. That's the design principle, and it's the one she's evangelizing across the industry.
She connects all of this back to her scientific background - published research on protein biochemistry, Alzheimer's disease, and Down's Syndrome from her MIT years - in an indirect but meaningful way. Science trained her to ask what the data actually says, not what you want it to say. In a category full of hype, that instinct is rare and useful.